Poster Title:
Summarization of Online Image Collections via Implicit Feedback
The availability of map interfaces and location-aware devices makes a growing amount of unstructured, geo-referenced information available on the Web. In particular, ten million tagged, geo-referenced photos are now available on Flickr, a photo-sharing website. We show how we analyze Flickr images to generate aggregate knowledge in the form of ``representative tags'' for arbitrary areas in the world. We display these tags on a map interface in an interactive web application along with images associated with each tag. We then use the aggregate user interactions with the tags and images to learn which images best describe the area shown on the map.